Dialysis

Cognitive Impairment and 7-Year Mortality in Dialysis Patients

Konstadina Griva, PhD,1,2 Jan Stygall, MSc,2 Matthew Hankins, MSc,3
Andrew Davenport, FRCP,4 Michael Harrison, FRCP,2 and Stanton P. Newman, PhD2
Background: Although dementia has predicted mortality in large dialysis cohorts, little is known
about the relationship between less pronounced cognitive deficits and mortality in patients with
end-stage renal disease. This study assessed whether cognitive impairment without dementia was an
independent predictor of 7-year survival in dialysis patients after controlling for other risk factors.
Study Design: Prospective single-cohort study.
Setting & Participants: 145 prevalent dialysis patients from 2 units in London, UK, were followed up
for 64.3 27.4 months and censored at the time of change to a different treatment.
Predictors: Cognitive impairment, defined as performance 1 standard deviation less than normative
values on 2 or more cognitive tests within a neurocognitive battery assessing attention/concentration,
memory, and psychomotor function domains. Depression, quality-of-life, and clinical measures also
were obtained.
Outcomes & Measurements: All-cause mortality was the primary outcome. Cox proportional hazard
models were used to assess the contribution of demographics and clinical and psychological measures
and cognitive impairment to mortality.
Results: 98 (67.6%) patients were cognitively impaired at baseline. At follow-up, 56 (38.6%) patients
had died, 29 of cardiac causes. Unadjusted Kaplan-Meier analysis showed higher mortality in cognitively impaired patients, in whom 7-year survival was 49% versus 83.2% in those with no cognitive
impairment (P 0.001). Mortality risk associated with cognitive impairment remained significant in
adjusted analysis controlling for sociodemographic, clinical, and psychological factors (adjusted HR,
2.53; 95% CI, 1.03-6.22; P 0.04).
Limitations: Small sample size and number of events.
Conclusions: Cognitive impairment is an independent predictor of mortality in dialysis patients.
Although the implications of early recognition and treatment of cognitive impairment for clinical
outcomes are unclear, these results suggest that patient management protocols should attempt to
ensure prevention of cognitive decline in addition to managing coexisting medical conditions.
Am J Kidney Dis 56:693-703. 2010 by the National Kidney Foundation, Inc.
INDEX WORDS: Cognition; mortality; dialysis; impairment.

Editorial, p. 615

ll-cause and cardiovascular mortality rates

for dialysis patients are high despite continual improvement in dialysis technology. This
may be a consequence of the increased number
of elderly patients and patients with concurrent
systemic diseases admitted to dialysis programs.1 Determinants of mortality in patients
with end-stage renal disease (ESRD) treated using hemodialysis include older age and comorbid
illnesses, such as diabetes mellitus, inflammation, and nutritional status.2-5 Psychosocial factors, such as depression,6,7 social support,8 and
quality of life (QoL), also have been associated
with mortality in dialysis patients.9-14
Cognitive dysfunction is a well-recognized
additional complication of ESRD, with recent

studies showing that deficits are apparent early in

the progression of kidney disease.15-17 Dialysis
patients have presented with impairments in cognitive abilities, such as poor memory, motor and

From the 1Department of Psychology, National University

of Singapore, Singapore; 2Unit of Behavioural Medicine, University College London and Health Services Research Group,
City University London; 3Institute of Psychiatry, Kings College
London; and 4Department of Nephrology, Royal Free and
Middlesex Hospital, London, UK.
Received January 16, 2010. Accepted in revised form July
1, 2010. Originally published online as doi:10.1053/j.ajkd.
2010.07.003 on August 27, 2010.
Address correspondence to Stanton P. Newman, PhD,
Health Services Research Group, City University London,
Northampton Square, London EC1V 0HB, UK. E-mail:
s.newman@ucl.ac.uk
2010 by the National Kidney Foundation, Inc.
0272-6386/10/5604-0013$36.00/0
doi:10.1053/j.ajkd.2010.07.003

American Journal of Kidney Diseases, Vol 56, No 4 (October), 2010: pp 693-703

693

694

Griva et al

psychomotor slowing, and attention deficits,18,19

although to date, the cause has not been established.19,20 Cognitive impairment has been linked
consistently with increased mortality in clinical
patient groups21-24 and community cohorts.25
This association is not restricted to elderly
samples.26 Lower cognitive ability in younger
adults and even children has been associated
with increased mortality several decades later.27
The limited research in patients with ESRD has
focused on the adverse clinical outcomes associated with a diagnosis of dementia28,29; however,
the association of less pronounced (minor) cognitive impairments, short of dementia, to mortality
has not been examined.
In this study, we assessed the prognostic significance of cognitive impairment (short of dementia) on 7-year survival in a cohort of clinically stable and adequately dialyzed dialysis
patients.

METHODS
Participants
After approval by the hospitals Institutional Review
Board and signed consent were obtained, 145 participants
were recruited between October 1998 and October 2000
from 2 dialysis units under the direction of University

Figure 1. Flow diagram of participant selection and

exclusion process. HD, hemodialysis; PD, peritoneal dialysis; TMT, Trail Making Test.

College and Royal Free Hospital Renal Department, London, UK (Fig 1). These included 52 patients on hospital
hemodialysis therapy, 25 on home hemodialysis therapy, and
68 on peritoneal dialysis therapy. Eligible participants had to
meet the following inclusion criteria: (1) 18 years or older;
(2) no history or clinically evident cerebrovascular disease;
(3) no major (uncorrected) visual or hearing impairments or
other sensory/motor impairments that prohibit them from
completing the scheduled assessments; (4) absence of acute
or chronic psychosis, documented depression, learning disabilities, and/or dementia (assessed from medical records,
abbreviated mental test scores,30 and staff review); (5)
currently stable, defined as not being acutely ill or hospitalized at the time of the assessment; (6) fluent in written and
spoken English; and (7) a minimum of 3 months on their
respective mode of treatment and dialysis techniques (eg,
the same dialysate or dialyzer if on hemodialysis therapy).

Measurements
Sociodemographic and Clinical Factors
Baseline data regarding sociodemographic factors (age,
sex, ethnicity, education, marital and employment status,
and household income), comorbid conditions (diabetes mellitus [as both kidney disease and a comorbid condition in
addition to another kidney disease], heart failure, ischemic
heart disease, peripheral vascular disease, malignancy, liver
disease, and lung disease), primary cause of ESRD, treatment history/dialysis vintage, laboratory values, residual
glomerular filtration rate (calculated as mean renal clearance
of urine and creatinine corrected for body surface), and
dialysis dose by equilibrated Kt/V were collected at patient
entry into the study.
Dialysis modality at baseline was defined as the type of
dialysis used at 4 weeks after enrollment in the study. All
forms of peritoneal dialysis (continuous ambulatory and
automated peritoneal dialysis) were combined as a single
category. Hospital hemodialysis and home hemodialysis
patients were collapsed into one group. Patients were considered to have switched dialysis technique when they changed
from one type of dialysis to another and continued to use the
new technique for at least 30 days.
Blood samples were obtained after the completion of each
neuropsychological testing session to avoid interference of
venipuncture pain with the participants performance. All
blood samples were delivered to the respective laboratories
within 2 hours of collection. Laboratory analyses consisted
of measurement of urea, creatinine, hemoglobin, potassium,
phosphate, calcium, and albumin.
Renal severity and comorbid conditions were assessed
using the ESRD Severity Index (ESRD-SI), a renal-specific
medical recordderived index completed by a nephrologist
familiar with the patient. The ESRD-SI measures severity of
illness as a function of 11 conditions (heart disease, cerebrovascular disease, peripheral vascular disease, peripheral neuropathy, bone disease, respiratory disease, visual impairment, access and dialysis events, autonomic neuropathy,
gastrointestinal disease and diabetes, and other).31
The modified Charlson Comorbidity Index (CCI) was used
to quantify comorbid illness. CCI scores were calculated using
the method previously described by Beddhu et al.32

Cognition and Mortality

695
Table 1. Description of Neuropsychological Tests

NP Tests

Cognitive Domain Assessed

Trail Making Test Part A

Attention/concentration

Trail Making Test Part B

Attention/concentration and
executive function

Symbol Digit Modality

Test

Attention/concentration

Rey Auditory Verbal

Learning Test
Benton Visual Retention
Test

Learning and verbal memory

Grooved Pegboard

Nonverbal memory/visual
perception and
visuocontructive abilities
Psychomotor speed/manual
dexterity

Test Description

Ability to efficiently connect

numbers in sequence
Ability to efficiently alternate
between sequential list of
numbers and letters
Visual scanning and rapid response
(written or oral) in substituting
symbols for numbers according
to learned code
Ability to retain and recall words
Ability to retain and replicate
designs immediately after 10-s
presentation
Ability to perform the task of picking
and placing pegs onto board
varying in orientation using 1
hand

Performance Measures

Time to completion (s)

Time to completion (s)

No. of correct substitutions

within 90 s (written or
oral)
Total of correct recalls
across 5 trials
No. correct; no. of errors

Time to completion in s
(dominant and
nondominant hands
measured separately)

Abbreviation: NP, neuropsychological.

Functional status of the patients was assessed by dialysis

staff using the Karnofsky scale.33

Neuropsychological Assessment
To identify and exclude dementia cases, medical records
were reviewed and health care professionals involved in
patients care screened the preliminary pool of eligible
patients with regard to cognitive and neuropsychiatric symptoms (eg, mental confusion, aggression, and self-care behaviors). The abbreviated mental test scores30 were subsequently used as a general cognitive screening measure for
dementia to determine eligibility for the study (cutoff score
of 8/9, depending on years of education).
Participants were also administered a battery of 6 validated neuropsychological tests assessing multiple cognitive
domains34 (Table 1): Trail Making Test Part A,35 Trail
Making Test Part B,35 Symbol Digit Modalities,36 Rey
Auditory Verbal Learning Memory Test,37 Benton Visual
Retention Test,38 and Grooved Pegboard.39 Testing was
conducted approximately 1 hour before the dialysis run or
clinic appointment (peritoneal dialysis patients only) at the
hospital or at home (home hemodialysis patients only).
Cognitive impairment was operationalized using results of
neuropsychological tests. Published norms for each test
were used to classify results by number of standard deviations (SDs) less than age- and, when available, sex- and
education-adjusted mean values.40 Cognitive impairment
was defined as performance of 1 SD less than the normative mean value on at least 2 measures, in line with the
Mayo41 and Diagnostic and Statistical Manual Disorders,
Fourth Edition, criteria for mild neurocognitive disorder.42
QoL was measured using the UK version of the 36-Item
Medical Outcome Study Short-Form Health Survey (SF36).43,44 The SF-36 is a generic multidimensional measure
of QoL that contains 8 subscales representing physical
functioning, social functioning, role limitations due to physi-

cal health problems, role limitations due to emotional problems, mental health, vitality, bodily pain, and general health
perceptions. These were combined into a Physical and a
Mental Component Summary score.45
Depressive symptoms were determined using the Beck
Depression Inventory-II (BDI),46 a well-validated 21-item
scale used extensively in ESRD populations. Scores can
range from 0-63, with a score 16 suggested as the optimal
cutoff for depression in patients with ESRD.47 In addition to
total BDI scores, the Cognitive Depression Index was computed using the subset of 14 cognitive/affective depression
items to control for the confounding contribution of somatic
symptoms of physical illness/uremia.48

Mortality
Patients were followed up for up to 7 years after enrollment. The primary end point was all-cause mortality. Mortality status (date and cause of death) was monitored through
the hospitals computerized medical records system, individual case records, and contact with primary renal physicians. Causes of death were defined according to the ERAEDTA (European Renal AssociationEuropean Dialysis and
Transplant Association) coding system.49 Cardiovascular
mortality was defined as myocardial ischemia and infarction, cardiac arrest, hypertensive and other causes of cardiac
failure, cerebrovascular accident, hemorrhage from ruptured
vascular aneurysm, or mesenteric infarction. Survival time
was calculated as the number of months from the baseline
assessment until one of the following: death, kidney transplant, loss to follow-up, transfer out of the facility, or end of
the study observation period (February 2006, at a maximum
of 89 months).

Statistical Analysis
Mean values and frequencies of the parameters were
compared using analysis of variance or 2 test, as appropri-

696

Griva et al

ate. Univariate analysis was conducted using Kaplan-Meier

survival tables for each (categorical/dichotomized) variable.
For continuous variables, univariate Cox regressions were
performed. To control for inflated type I error rate, variables
were considered significant predictors for P 0.005.
Significant univariate predictors were entered into a multivariate Cox regression model in which the relationship
between cognitive functioning and survival was examined
while simultaneously controlling for sociodemographic, clinical, and psychological factors. A forward conditional stepwise entry method was used to generate hazard ratios (HRs)
with 95% confidence limits.
Patients were followed up until classified as dead or
censored. Reasons for censoring included loss to follow-up
(ie, transplant or transfer out of the facility) or end of
follow-up in February 2006. Analysis was carried out using
an intention-to-treat approach in which death was assigned
to a patients initial dialysis modality regardless of a change
in dialysis treatment during the course of follow-up.

RESULTS
Characteristics of Study Population
The total study sample surveyed included 145
participants, an 88.4% overall enrollment rate
(mean follow up, 64.3 27.4 [SD] months).
Cognitive dysfunction (scores in 2 neuropsychological tests 1 SD lower than normative values) was detected in 98 (67.6%) of the patients
assessed at baseline. Impairment was evident
across multiple domains, with the highest prevaTable 2. Frequency of Cognitive Impairment Across
Neuropsychological Tests

NP Test

None

Mild
(1-1.99 SDs
< norms)

TMT-A (%)
TMT-B (%)
SDMT-W (%)
SDMT-O (%)
RAVLT (%)
BVRT-C (%)
BVRT-E (%)
GP-D (%)
GP-ND (%)

78.6
51.0
67.9
57.6
51.0
39.3
43.4
57.2
52.1

15.9
24.2
25.5
37.6
26.2
35.2
29.6
24.0
27.8

Moderate
(2-2.99 SDs
< norms)

5.5
24.8
7.6
10.3
22.8
25.5
27.0
18.8
20.1

Abbreviations: BVRT-C, Benton Visual Retention Test

correct designs; BVRT-E, Benton Visual Retention Test
errors; GP-D, Grooved Pegboard dominant hand; GP-ND,
Grooved Pegboard nondominant hand; norm, normative
value; NP, neuropsychological; RAVLT-T, Rey Auditory
Verbal Learning Test total word recall at trials 1-5; SD,
standard deviation; SDMT-O, Symbol Digit Modality Test
oral administration; SDMT-W, Symbol Digit Modality Test
written administration; TMT-A, Trail Making Test Part A;
TMT-B, Trail Making Test Part B.

lence in tests of memory and executive function

and lowest in simple tests of attention (Table 2).
The most frequently occurring pattern of these
cognitive deficits involved the 3 domains of
memory, executive function, and motor skills.
Those with cognitive impairment were older,
were less likely to be employed or perceiving
themselves as being able to work, had lower
income, had a worse clinical profile indexed
using lower Karnofsky and higher ESRD severity scores, had a greater prevalence of diabetes
and cardiovascular disease, and had lower clearance and lower albumin concentrations than those
without cognitive impairment (Tables 3 and 4).
Sex and clinical characteristics, namely dialysis modality; dialysis vintage; time on renal
replacement therapy; hemoglobin, urea, and creatinine levels; dialysis adequacy; and CCI scores,
were not associated with cognitive impairment.
Mortality Rates
During the study period, 56 (38.6%) patients
died (Table 5). Other reasons for censoring included transfer to another unit (7 patients) and
transplant (59 patients). Of 56 deaths, 50 (51%)
and 6 (12.8%) were patients with and without
cognitive impairment, respectively.
The main cause of death was cardiovascular
disease (29 patients). These included myocardial
ischemia and infarction (10 patients), heart failure (12 patients), and cardiac arrest (7 patients).
Other causes of death included cerebrovascular
accident (3 patients), infection (11 patients), withdrawal from treatment (6 patients), malignancies
(2 patients), miscellaneous (namely, acute liver
failure and accident related to ESRD treatment; 2
patients), and unknown (3 patients).
There were no significant differences in distributions of causes of death between patients (cardiac vs all other) with and without cognitive
impairment at baseline (data not shown).
Survival Analysis
Univariate analysis using log-rank tests to
compare Kaplan-Meier survival curves or univariate Cox regressions (as appropriate) indicated
that age, employment, ESRD-SI, diabetes, cardiovascular disease, vascular disease, physical QoL
(Physical Component Summary score), depression, and cognitive impairment were significantly associated with survival (P 0.001).

Cognition and Mortality

697

Table 3. Baseline Characteristics of Patients With and Without Cognitive Impairment

Total (N 145)

Cognitive
Impairment (n 98)

No Cognitive
Impairment (n 47)

Sociodemographic
Age (y)
Age at diagnosis (y)
Education (y)
Women (%)
Ethnicity (% white)
Married (%)
Employed (%)
Able to work (%)

50.12 14.26
44.52 16.01
12.37 5.41
35.2
64.1
60.0
35.9
43.4

51.81 14.05
46.61 15.66
12.29 6.09
37.8
59.2
62.2
24.5
30.6

46.6 14.2
40.17 16.01
12.53 3.68
29.8
74.5
55.3
59.6
70.2

0.04
0.02
0.8
0.4
0.6
0.4
0.001
0.04

Income ()
0-10,000 (%)
10,001-20,000 (%)
20,001-30,000 (%)
30,001 (%)

58.6
24.8
9.7
6.9

68.4
24.5
4.1
3.1

38.3
25.5
21.3
14.9

0.001
0.5
0.001
0.001
0.7

Dialysis modality (% PD)

46.9

48.0

44.7

Time on RRT (mo)

65.36 74.32

58.78 74.43

79.09 72.94

0.1

Time on dialysis (mo)

49.80 58.32

45.95 59.56

57.85 55.42

0.1

Laboratory data
Urea (mg/dL)
Potassium (mmol/L)
Creatinine (mg/dL)
Calcium (mg/dL)
Phosphate (mg/dL)
Albumin (g/dL)
Hemoglobin (g/dL)

62.04 15.32
4.98 0.92
10.53 2.71
9.66 0.88
5.33 1.73
3.79 0.23
11.03 1.46

61.93 15.15
4.97 0.92
10.24 2.7
9.50 0.88
5.23 1.55
3.72 0.45
11.00 1.46

62.24 15.85
5.00 0.92
11.15 2.66
9.98 0.84
5.61 2.01
3.93 0.35
11.11 1.47

0.9
0.8
0.7
0.06
0.2
0.005
0.7

Clinical measures
Kt/Vurea/wk (HD)
Kt/Vurea/wk (PD)
Residual GFR (mL/min/1.73 m2)
Karnofsky score

1.69 0.23
1.98 0.41
4.2 3.1
88.04 11.28

1.63 0.23
1.87 0.38
4.3 3.1
85.46 11.9

1.79 0.21
2.14 0.40
3.9 2.8
93.48 7.37

0.08
0.01
0.4
0.001

Comorbidity
ESRD-SI
CCI score
Hypertension (%)
Cardiac diseasea (%)
Vascular diseaseb (%)
Diabetes mellitus (%)

11.16 9.47
4.54 2.36
91.7
39.6
54.4
17.2

13.18 9.63
4.55 2.5
92.9
46.9
58.1
25.5

6.96 7.64
4.51 2.08
89.4
25.5
46.8
0.0

0.001
0.9
0.5
0.001
0.004
0.001

13.8
13.8
9.7
11.0
10.3
10.4
13.8
13.1
4.1

12.2
8.2
7.1
16.3
10.2
11.2
12.3
17.4
5.1

17.0
25.5
14.9
0.0
10.6
8.5
17.0
4.4
2.1

0.08
0.8
0.5
0.001
0.9
0.9
0.4
0.7
0.6

Primary kidney disease

Glomerulonephritis (%)
Polycystic renal disease (%)
Reflux nephropathy (%)
Diabetic nephropathy (%)
Hypertension (%)
Renal malformation (%)
Other nephropathies (%)
Miscellaneous (%)
Uncertain (%)

Note: Values are expressed as mean standard deviation or percentage. Abbreviations: CCI, Charlson Comorbidity
Index; ESRD-SI, End-Stage Renal Disease Severity Index; GFR, glomerular filtration rate; HD, hemodialysis; PD, peritoneal
dialysis; RRT, renal replacement therapy.
a
History of coronary disease or congestive heart failure.
b
History of angina, myocardial infarction, transient ischemic attack, symptomatic peripheral vascular disease, or arterial
bypass surgery.

698

Griva et al
Table 4. Neuropsychological Test Scores of Patients With and Without Cognitive Impairment
Total (N 145)

Cognitive Impairment
(n 98)

No Cognitive Impairment
(n 47)

NP test scores
TMT-Aa
TMT-Ba
SDMT-Wb
SDMT-Ob
RAVLT-Tb
GP-Da
GP-NDa
BVRT-Cb
BVRT-Ec

52.21 32.43
98.57 49.39
41.10 12.59
45.39 13.73
39.03 10.71
91.02 31.97
1.49 0.50
4.92 2.16
8.56 5.02

61.86 34.85
114.11 52.14
36.15 10.86
39.69 11.55
36.06 10.66
100.56 34.31
1.68 0.47
4.21 2.00
10.10 5.08

32.09 10.94
66.18 17.84
51.43 9.29
57.28 9.77
45.21 7.85
70.69 9.47
1.09 0.28
6.40 1.69
5.34 2.98

0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001

Psychological measures
BDI
CDI
PCS score
MCS score

12.03 8.31
7.29 5.79
28.87 13.19
47.24 10.37

13.90 8.84
8.40 6.20
25.43 12.15
46.59 10.77

8.13 5.34
4.98 3.98
36.04 12.49
48.57 9.45

0.001
0.001
0.01
0.3

Note: Values are expressed as mean standard deviation.

Abbreviations: BDI, Beck Depression Inventory; BVRT-C, Benton Visual Retention Test correct designs; BVRT-E, Benton
Visual Retention Test errors; CDI, Cognitive Depression Index; GP-D, Grooved Pegboard dominant hand; GP-ND, Grooved
Pegboard nondominant hand; MCS, Mental Component Summary; NP, neuropsychological; PCS, Physical Component
Summary; RAVLT-T, Rey Auditory Verbal Learning Test total word recall at trials 1-5; SDMT-O, Symbol Digit Modality Test
oral administration; SDMT-W, Symbol Digit Modality Test written administration; TMT-A, Trail Making Test Part A; TMT-B,
Trail Making Test Part B.
a
Time to completion in seconds.
b
Number correct.
c
Number of errors.

Survival rates were 49% for patients with

cognitive impairment and 83.2% for those with
no impairment (log-rank 2 19.61; P 0.001).
Figure 2 shows unadjusted survival curves for
cognitive impairment during the study followup. In the unadjusted Cox proportional hazards
model, cognitive impairment was associated with
mortality, with an HR of 5.16 (95% confidence
interval, 2.21-12.06; P 0.001).
Treatment modality, sex, education level, income, ethnicity, marital status, primary cause of
kidney disease, Karnofsky scores, CCI score,
treatment modality, dialysis center, and dialysis
characteristics, including vintage, residual glomerular filtration rate, Kt/V, and laboratory biochemical values (baseline urea, creatinine, albu-

min, hemoglobin, calcium, potassium, and

phosphate) were not associated significantly with
mortality in univariate analyses.
Multivariate Analysis
All variables significantly associated with survival in unadjusted univariate analyses were included in the adjusted multivariate adjusted Cox
regression model. Potential predictors were
grouped a priori into 4 conceptually based blocks,
and each block of variables was added sequentially into the model as follows: (1) demographic
variables: age and employment; (2) medical risk
factors: ESRD-SI, diabetes mellitus, cardiovascular disease, and vascular disease; (3) psychologi-

Table 5. Outcomes (survival) in Patients With and Without Cognitive Impairment

Count of deaths (% of deaths)

Survival time (mo)

Total (N 145)

Cognitive
Impairment (n 98)

No Cognitive
Impairment (n 47)

56 (38.6)
64.33 27.38

48 (49.0)
58.29 29.11

6 (12.8)
76.91 17.94

0.001
0.001

Cognition and Mortality

699

1.0

Cumulativ
ve Survival

0.8

0.6
Neuropsychological
Impairment

0.4

No impairment
Impairment
No impairmentcensored
p
Impairment-censored

0.2

0.0
0

4
Time (y)

Figure 2. Unadjusted survival curves by cognitive functioning at baseline to the end of follow-up (n 145).

cal risk factors: physical QoL and depression;

and (4) cognitive functioning/impairment.
Results indicated that cognitive impairment
was independently predictive of death (adjusted HR, 2.53; 95% confidence interval, 1.036.22) even after stepwise adjustment for demographic, medical, and psychological factors.
Other significant predictors were ESRD severity, diabetes, and Physical Component Summary score (Table 6).
Associations of cognitive impairment to mortality were similar when the analysis was repeated without censoring patients at transplant.
Similar results also were obtained when survival
time for each patient was determined by both the
number of months between the first dialysis and
the end-of-study observation period or date of
death and when depression scores were considered as continuous variables (data not shown).

DISCUSSION
In this prospective study, we examined mortality
in a cohort of 145 prevalent dialysis patients and
performed a risk-factor analysis. After a 7-year
observation period, 38.6% of patients had died. The
observed mortality rate in this patient cohort was
lower than that reported in other registries and
study populations in the United States and Europe50-51 and may reflect sample differences. The
mean age and percentage of patients with diabetes
was lower in this study than in equivalent cohorts.50,51 Cardiovascular causes accounted for the

largest percentage of deaths in accordance with

previous research.52,53 Despite excluding frank dementia in the recruiting of this dialysis cohort, a
very high proportion of participants had at least
mild cognitive impairment, in line with reports in
older dialysis cohorts.54 Rates of cognitive impairment contrast unfavorably with reported rates in
the general older population.55
Cognitive functioning emerged as a unique
prognostic factor in predicting survival in the
dialysis cohort, extending previous findings on
dialysis patients with diagnosed dementia28,29
and studies in other clinical populations24,56 and
community cohorts.26 Cognitive impairment in 2
or more tests at baseline was associated with
nearly 3 times greater risk of death 7 years later,
even after extensive case-mix adjustments. This
is the first report of higher mortality rates in
dialysis patients with cognitive impairments in
the absence of dementia.
Several possible explanations may account for
this finding. One possibility is that cognitive
impairments may lead to increased mortality by
impeding treatment adherence.18,57 Optimal selfmanagement for dialysis patients is complex and
requires patients to understand, remember, and
reason to follow treatment advice and modify
behavior (eg, in relation to diet, fluid intake, or
medication). Cognitive deficits may result in
suboptimal care because of difficulty following
treatment recommendations and adhering to and
Table 6. Stepwise Cox Regression Model for
All-Cause Mortality
B

Wald

Age (/y)
0.01 0.57
Employment
0.10 0.06
ESRD-SI score
0.44 6.01
(/1-point increase)
Diabetes
0.72 6.19
PCS score (/1-point 0.45 5.81
increase)
NP impairment
0.93 4.01

HR

95% CI

0.5 1.01 0.99-1.04

0.8 1.10 0.51-2.41
0.01 1.56 1.10-2.14
0.02 2.06 1.13-3.74
0.01 0.64 0.44-0.92
0.04 2.53 1.03-6.23

Note: All variables are standardized to a standard deviation of 1 except for age and binary predictors (employment,
diabetes, cardiovascular disease, depression, and neuropsychological impairment). Values are based on final step
of regression.
Abbreviations: CI, confidence interval; ESRD-SI, EndStage Renal Disease Severity Index; HR, hazard ratio; NP,
neuropsychological; PCS, Physical Component Summary.

700

managing multiple medications, as evidenced in

other clinical populations.22,58
An alternative explanation is that impaired
cognition may be a surrogate marker of global
decline in health or disease progression. Population-based studies have shown that general ill
health is associated with cognitive deficits.59 It
therefore is plausible that poor cognitive ability
in dialysis patients is related to mortality because
cognitive function is a sensitive indicator of
general physical deterioration. It is possible that
cognitive deterioration may reflect a specific
disease process. Patients with chronic kidney
disease show high levels of vascular disease and
related processes.60 Vascular stiffness has been
shown to start at early stages of chronic kidney
disease with morphologic alterations and structural changes manifested even in pediatric patients with much lower exposure to classical
risk factors for atherosclerosis, such as diabetes
or hyperlipidemia.61 Cognitive dysfunction may
reflect this vascular pathologic state.62,63 This is
consistent with evidence showing that impaired
neuropsychological scores are associated with
measures of cerebrovascular damage, such as
white matter hyperintensities, in both hemodialysis64-66 and peritoneal dialysis patients67 and
predict incident cardiovascular disease.68,69 It
therefore is conceivable that the reported associations between cognition and mortality are the
result of subclinical atherosclerosis or ischemic
changes caused by vascular disease and/or smallvessel disease not fully accounted for in these
analyses. In the analyses, adjustments for severity of ESRD and comorbid conditions, including
established vascular risk factors and physical
QoL, did not attenuate the mortality risk associated with poor cognition.
The expected multivariate associations of age,
serum albumin level, hemoglobin concentration,
dose of dialysis administered, and depression
with mortality were not statistically significant in
this study, although the trends and observed
univariate associations were similar to those in
previous studies.4-7 The lack of association between the clinical indices (Kt/V and albumin and
hemoglobin levels) and survival may reflect the
marked reduction in variability and improved
care for anemia and adequacy of dialysis in our
sample because all patients in this dialysis cohort
achieved clinical targets. For instance, Lowrie

Griva et al

and Lew70 showed that the risk of death increases exponentially as serum albumin concentration decreases from 4.0 to 2.5 g/dL. Therefore,
the notion that these clinical markers are independent predictors may be more likely to be true at
lower serum levels that do not meet clinical
targets.
In this sample, we did not confirm previous
findings of an increased risk of death associated
with depression.7,71,72 This may relate to differences in follow-up duration and assessment procedures. In this study, depression was measured
only at baseline without subsequent longitudinal
periodic assessment. Depressive mood can be
labile and it may be unreasonable to expect
depressive mood at a single measurement to be
strongly predictive of outcomes several years
later. Chronic persistent depressive mood with
multiple assessments during the course of the
study, rather than a single baseline measurement,
may be more likely to be associated with mortality.73,74
Although the findings of this study are novel,
several limitations must be acknowledged. In
comparison to other studies, sample size and
resulting number of deaths were small. This did
not allow us to perform separate analyses for
different dialysis modalities or various other subgroups. Thus, although the magnitude of the HR
is large, issues related to study power limitations
highlight the importance of revisiting the question with larger samples.
Second, the study sample comprised prevalent
dialysis patients. Given the excess early mortality in incident patients with ESRD,75 our cohort
likely represents a healthier patient population
who survived the initial dialysis period. In addition, because participants were selected on the
basis of strict eligibility criteria, results may not
be generalizable to the dialysis population as a
whole. The prevalence of cognitive impairment
and mortality may have been underestimated
because patients with a history of stroke were
excluded.
Another potential limitation is that the screening to exclude patients with dementia was based
on medical history and abbreviated test scores
with no formal cognitive/neurologic diagnostic
evaluation. Cases of (early) dementia may have
gone undetected and thereby biased the prevalence estimate of cognitive impairment. How-

Cognition and Mortality

ever, the absence of severe cognitive impairment

in neuropsychological tests (ie, performance 3
SDs less than normative values) coupled with
findings of preserved daily functioning and low
mean age suggest that our sample most likely
was dementia free at the time of assessment.
Third, this report does not include timedependent analyses. Cognitive functioning was
measured only once at baseline; thus, we do not
know whether cognitive function improved or
declined during follow-up. In all probability,
cognitive functioning may have deteriorated further, possibly progressing to dementia in some
cases,76 and more patients may have developed
cognitive dysfunction during the 7-year followup, but these effects were not assessed.
Finally, even in longitudinal data such as these,
issues of causality cannot be answered definitively. Although the analyses presented here
showed that compromised cognitive functioning
preceding death by as much as 7 years was
strongly associated with higher risk of mortality,
we cannot conclude that high levels of cognitive
abilities cause dialysis patients to live longer. By
controlling for illness severity markers, comorbid conditions, and other well-being indices (QoL
and depression) at the time of data collection, we
were able to show that higher levels of cognition
are protective well in advance of death and not
simply an indicator of better health status. This
suggests that the association between cognitive
function and mortality may reflect processes different from those underlying a simple relation
between comorbid conditions and mortality.
Despite its limitations, this study has important clinical implications. Previously, mild cognitive impairment has been overlooked as a potentially important factor in the clinical management
of dialysis patients. Data presented here indicate
that cognitive functioning can serve as a sensitive indicator of subsequent mortality. These data
suggest that it may be valuable to include a
cognitive assessment in the routine evaluation of
patients on dialysis therapy. Various cognitive
interventions successful in other populations77,78
assuming causality could potentially improve or
halt decrements in cognitive functioning associated with ESRD, thus decreasing the risk of
mortality, possibly through improved adherence.
In conclusion, this is the first demonstration of
a strong association between poor cognitive func-

701

tioning and mortality in a dialysis sample after

extensive adjustment for case-mix and other
known (clinical and psychosocial) risk factors.
Replication of this finding in other renal patient
samples is necessary. Future work should address the questions regarding predictors of cognitive functioning in patients with ESRD and explore the potential mechanisms through which
cognition impacts on mortality.

ACKNOWLEDGEMENTS
The authors thank the renal staff for their support and the
patients for giving their time to complete the research
protocol.
Support: This research was supported by grants from
Alexandros Onassis Foundation (to Dr Griva) and the R.L.
Weston Institute for Neurological Studies of the UCL Medical School, both of which are gratefully acknowledged.
Financial Disclosure: The authors declare that they have
no relevant financial interests.

REFERENCES
1. Jager KJ, van Dijk PC, Dekker FW, et al. The epidemic
of aging in renal replacement therapy: an update on elderly
patients and their outcomes. Clin Nephrol. 2003;60(5):352360.
2. Goodkin DA, Bragg-Gresham JL, Koenig KG, et al.
Association of comorbid conditions and mortality in HD
patients in Europe, Japan, and the United States: the Dialysis
Outcomes and Practice Patterns Study (DOPPS). J Am Soc
Nephrol. 2003;14(12):3270-3277.
3. Foley RN, Parfrey PS, Sarank MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J
Kidney Dis. 1998;32(suppl 3):S112-119.
4. Kakiya R, Shoji T, Tsujimoto Y, et al. Body fat mass
and lean mass as predictors of survival in hemodialysis
patients. Kidney Int. 2006;70(3):549-556.
5. Cooper BA, Penne EL, Bartlett LH, et al. Protein
malnutrition and hypoalbuminemia as predictors of vascular
events and mortality in ESRD. Am J Kidney Dis. 2004;43(1):
61-66.
6. Kimmel P. Depression in patients with chronic renal
disease: what we know and what we need to know. Psychosom Res. 2002;53(4):951-956.
7. Kimmel PL, Peterson RA, Weihs KL, et al. Multiple
measurements of depression predict mortality in a longitudinal study of chronic hemodialysis outpatients. Kidney Int.
2000;57(5):2093-2098.
8. Spinale J, Cohen SD, Khetpal P, et al. Spirituality,
social support, and survival in hemodialysis patients. Clin
J Am Soc Nephrol. 2008;3(6):1620-1627.
9. DeOreo PB. Hemodialysis patient-assessed functional
health status predicts continued survival, hospitalization,
and dialysis-attendance compliance. Am J Kidney Dis. 1997;
30(2):204-212.
10. Kalantar-Zadeh K, Kopple JD, Block G, et al. Association among SF36 quality of life measures and nutrition,

702
hospitalization, and mortality in hemodialysis. J Am Soc
Nephrol. 2001;12(12):2797-2806.
11. Mapes DL, Lopes AA, Satayathum S, et al. Healthrelated quality of life as a predictor of mortality and hospitalization: the Dialysis Outcomes and Practice Patterns Study
(DOPPS). Kidney Int. 2003;64(1):339-349.
12. Lopez RK, Garcia Lopez FJ, de Alvaro MF, et al.
Perceived mental health at the start of dialysis as a predictor
of morbidity and mortality in patients with end-stage renal
disease (CALVIDIA Study). Nephrol Dial Transplant. 2004;
19(9):2347-2353.
13. Lowrie EG, Curtin RB, LePain N, et al. Medical
Outcomes Study Short Form-36: a consistent and powerful
predictor of morbidity and mortality in dialysis patients.
Am J Kidney Dis. 2003;41(6):1286-1292.
14. Paniagua R, Amato D, Vonesh E, et al. Mexican
Nephrology Collaborative Study Group. Health related quality of life predicts outcomes but is not affected by peritoneal
clearance: the ADEMEX trial. Kidney Int. 2005;67(3):
1093-2104.
15. Buchman AS, Tanne D, Boyle PA, Shah RC, Leurgans SE, Bennett DA. Kidney function is associated with the
rate of cognitive decline in the elderly. Neurology. 2009;
73(12):920-927.
16. Elias MF, Elias PK, Seliger SL, Narsipur SS, Dore
GA, Robbins MA. Chronic kidney disease, creatinine and
cognitive functioning. Nephrol Dial Transplant. 2009;24(8):
2446-2452.
17. Kurella TM, Wadley V, Yaffe K, et al. Kidney function and cognitive impairment in US adults: the Reasons for
Geographic and Racial Differences in Stroke (REGARDS)
Study. Am J Kidney Dis. 2008;52(2):227-234.
18. Pereira AA, Weiner DE, Scott T, et al. Cognitive
function in dialysis participants. Am J Kidney Dis. 2005;
45(3):448-462.
19. Murray AM, Tupper DE, Knopman DS, et al. Cognitive impairment in hemodialysis patients is common. Neurology. 2006;67(2):216-223.
20. Griva K, Newman S, Harrison M, et al. Acute neuropsychological changes in hemodialysis and peritoneal dialysis patients. Health Psychol. 2003;22(6):570-578.
21. Anstey KJ, Mack HA, von Sanden C. The relationship between cognition and mortality in patients with stroke,
coronary heart disease, or cancer. Eur Psychol. 2006;11(3):
182-195.
22. McGuire L, Ford E, Ajani U. The impact of cognitive
functioning on mortality and the development of functional
disability in older adults with diabetes: the second longitudinal study on aging. BMC Geriatr. 2006;6:8.
23. Zuccal G, Pedone C, Cesari M, et al. The effects of
cognitive impairment on mortality among hospitalized patients with heart failure. Am J Med. 2004;115(2):97-103.
24. McLennan SN, Pearson SA, Cameron J, et al. Prognostic importance of cognitive impairment in chronic heart
failure patients: does specialist management make a difference? Eur J Heart Fail. 2006;8(5):494-501.
25. Liu IY, LaCroix AZ, White LR, et al. Cognitive
impairment and mortality: a study of possible confounders.
Am J Epidemiol. 1990;132(1):136-143.

Griva et al
26. Shipley BA, Der G, Taylor MD, et al. Cognition and
all-cause mortality across the entire adult age range: Health
and Lifestyle Survey. Psychosom Med. 2006;68(1):17-24.
27. Osler M, Andersen AM, Due P, et al. Socioeconomic
position in early life, birth weight, childhood cognitive
function, and adult mortality: a longitudinal study of Danish
men born in 1953. J Epidemiol Community Health. 2003;
57(9):681-686.
28. Kurrela M, Mapes DC, Port FK, et al. Correlates and
outcomes of dementia among dialysis patients: the Dialysis
Outcomes and Practice Patterns Study. Nephrol Dial Transplant. 2006;21(9):2543-2548.
29. Rakowski DA, Caillard S, Agodoa LY, Abbott KC.
Dementia as a predictor of mortality in dialysis patients. Clin
J Am Soc Nephrol. 2006;1(5):1000-1005.
30. Hodkinson HM. Evaluation of a mental test score for
assessment of mental impairment in the elderly. Age Ageing.
1972;1(4):233-238.
31. Craven J, Littlefield C, Rodin G, et al. The Endstage
Renal Disease Severity Index (ESRD-SI). Psychol Med.
1991;21(1):237-243.
32. Beddhu S, Bruns FJ, Saul M, Seddon P, Zeidel ML. A
simple comorbidity scale predicts clinical outcomes and
costs in dialysis patients. Am J Med. 2000;108(8):609-613.
33. Karnofsky DA, Burcherval JH. The clinical evaluation
of chemotherapeutic agents in cancer. In: Macleod CM, ed.
Evaluation of Chemotherapeutic Agents in Cancer. New York,
NY: Columbia University Press; 1949:191-205.
34. Lezak MD. Neuropsychological Assessment. Oxford,
UK: Oxford University Press; 2004.
35. Reitan RM, Wolfson D. The HalsteadReitan Neuropsychological Test Battery: Theory and Clinical Interpretation. Tucson, AZ: Neuropsychology Press; 1993.
36. Smith A. Symbol Digit Modalities Test Manual. Los
Angeles, CA: Western Psychological Services; 1973.
37. Rey A. Lexamen clinique en psychologie. Paris,
France: Presses Universitaires de France; 1964.
38. Benton AL. The Revised Benton Visual Retention
Test. New York, NY: Psychological Corp; 1974.
39. Matthews CG, Klove H. Instruction Manual for the
Adult Neuropsychology Test Battery. Madison, WI: University of Wisconsin Medical School; 1964.
40. Mitrushina MN, Boone KB, DElia LF. Handbook of
Normative Data for Neuropsychological Assessment. Oxford, UK: Oxford University Press; 1999.
41. Knopman DS, Boeve BF, Petersen RC. Essentials of
the proper diagnoses of mild cognitive impairment, dementia, and major subtypes of dementia. Mayo Clin Proc.
2003;78(10):1290-1308.
42. American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders. 4th ed. Washington,
DC: American Psychiatric Association, 1994.
43. Ware JE, Sherbourne CD. The MOS 36-Item ShortForm Health Survey (SF-36). Med Care. 1992;30(6):473483.
44. Jenkinson C, Stewart-Brown S, Petersen S, et al.
Assessment of the SF-36 version 2 in the United Kingdom. J
Epidemiol Community Health. 1999;53(10):46-50.
45. Ware JE, Kosinski M, Keller SK. Physical and Mental Health Summary Scales. A Users Manual. Boston, MA:
The Health Institute; 1994.

Cognition and Mortality

46. Beck AT, Steer RA, Garbin MG. Psychometric properties of the Beck Depression Inventory: twenty-five years
of evaluation. Clin Psychol Rev. 1988;8(1):77-100.
47. Watnick S, Wang PL, Demadura T, et al. Validation of
2 depression screening tools in dialysis patients. Am J
Kidney Dis. 2005;46(5):919-924.
48. Sacks CR, Peterson RA, Kimmel PL. Perception of
illness and depression in chronic renal disease. Am J Kidney
Dis. 1990;15(1):31-39.
49. Van Dijk PC, Jager KJ, de Charro F, et al. Renal
replacement therapy in Europe: the results of a collaborative
effort by the ERA-EDTA registry and six national or regional registries. Nephrol Dial Transplant. 2001;16(6):11201129.
50. US Renal Data System. USRDS 2008 Annual Data
Report: Atlas of Chronic Kidney Disease and End-Stage
Renal Disease in the United States. Bethesda, MD: National
Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2008.
51. ERA-EDTA Registry. ERA-EDTA Registry Annual
Report 2007. Amsterdam, the Netherlands: Academic Medical Center, Department of Medical Informatics; 2009.
52. Gill JS, Ma I, Landsberg D, et al. Cardiovascular
events and investigation in patients who are awaiting cadaveric kidney transplantation. J Am Soc Nephrol. 2005;16(3):
808-816.
53. Go AS, Chertow GM, Fan D, et al. Chronic kidney
disease and the risks of death, cardiovascular events, and
hospitalization. N Engl J Med. 2004;351(1):1296-1305.
54. Murray AM, Tupper DE, Knopman DS, et al. Cognitive impairment in hemodialysis patients is common. Neurology. 2006;67(2):216-223.
55. Lopez OL, Jagust WJ, DeKosky ST, et al. Prevalence
and classification of mild cognitive impairment in the Cardiovascular Health Study Cognition Study: part 1. Arch Neurol.
2003;60(10):1385-1389.
56. Oksala N, Jokinen H, Melkas S, et al. Cognitive
impairment predicts poststroke death in long-term follow
up. J Neurol Neurosurg Psychiatry. 2009;80(11):1230-1235.
57. Kimmel PL, Peterson RA, Weihs KL, et al. Psychosocial factors, behavioral compliance and survival in urban
hemodialysis patients. Kidney Int. 1998;54(1):245-254.
58. Hinkin CH, Hardy DJ, Mason KI, et al. Medication
adherence among HIV adults: effects of cognitive dysfunction and regimen complexity. Neurology. 2002;59(12):944950.
59. Tilvis RS, Khnen-Vre MH, Jolkkonen J, et al.
Predictors of cognitive decline and mortality of aged people
over a 10-year period. J Gerontol A Biol Sci Med Sci.
2004;59(3):268-274.
60. Ikram MA, van Oijen M, de Jong FJ, et al. Unrecognized myocardial infarction in relation to risk of dementia
and cerebral small vessel disease. Stroke. 2008;39(5):14211426.
61. Litwin M, Whl E, Jourdan C, et al. Altered morphologic properties of large arteries in children with chronic
renal failure and after renal transplantation. J Am Soc Nephrol. 2005;16(5):1494-1500.
62. Pecoits-Filho R, Lindholm B, Stenvinkel P. The malnutrition, inflammation and atherosclerosis (MIA) syn-

703
dromethe heart of the matter. Nephrol Dial Transplant.
2002;17(suppl 18):S28-31.
63. Murray AM. The brain and the kidney connection: a
model of accelerated vascular cognitive impairment. Neurology. 2009;73(12):916-917.
64. Dufouil C, Alperovitch A, Tzourio C. Influence of
education on the relationship between white matter lesions
and cognition. Neurology. 2003;60(5):831-836.
65. Kuller LH, Shemanski L, Manolio T, et al. Relationship between ApoE, MRI findings, and cognitive function in
the Cardiovascular Health Study. Stroke. 1998;29(2):388398.
66. Desmond DW. Cognition and white matter lesions.
Cerebrovasc Dis. 2002;13(suppl 2):S53-S57.
67. Kim CD, Lee HJ, Kim DJ, et al. High prevalence of
leukoaraiosis in cerebral magnetic resonance images of
patients on peritoneal dialysis. Am J Kidney Dis. 2007;50(1):
98-107.
68. Elkins JS, Knopman DS, Yaffe K, et al. Cognitive
function predicts first-time stroke and heart disease. Neurology. 2005;64(10):1750-1755.
69. De Fries, T Avendao M, Glymour MM. Level and
change in cognitive test scores predict risk of first stroke.
J Am Geriatr Soc. 2009;57(3):499-505.
70. Lowrie EG, Lew NL. Death risk in hemodialysis
patients: the predictive value of commonly measured variables and an evaluation of death rate differences between
facilities. Am J Kidney Dis. 1990;15(5):458-482.
71. Lopes AA, Bragg J, Young E, et al. Depression as a
predictor of mortality and hospitalization among hemodialysis patients in the United States and Europe. Kidney Int.
2002;62(1):199-207.
72. Hedayati SS, Bosworth HB, Briley LP, et al. Death or
hospitalization of patients on chronic hemodialysis is associated with a physician-based diagnosis of depression. Kidney
Int. 2008;74(7):930-936.
73. Kimmel PL, Peterson RA, Weihs KL, et al. Multiple
measurements of depression predict mortality in a longitudinal study of chronic hemodialysis outpatients. Kidney Int.
2000;57(5):2093-2098.
74. Boulware LE, Liu Y, Fink NE, et al. Temporal relation among depression symptoms, cardiovascular disease
events, and mortality in end-stage renal disease: contribution
of reverse causality. Clin J Am Soc Nephrol. 2006;1(3):496504.
75. Bradbury BD, Fissell RB, Albert JM, et al. Predictors
of early mortality among incident vs hemodialysis patients
in the Dialysis Outcomes and Practice Patterns Study
(DOPPS). Clin J Am Soc Nephrol. 2007;2(1):89-99.
76. Tuokko H, Frerichs R, Graham J, et al. Five-year
follow-up of cognitive impairment with no dementia. Arch
Neurol. 2003;60(4):577-582.
77. Camp CJ, Foss JW, Stevens AB, et al. Memory
training in normal and demented elderly populations: the
E-I-E-I-O model. Exp Aging Res. 1993;19(3):277-290.
78. Willis SL, Nesselroade CS. Long-term effect of fluid
ability training in old-old age. Dev Psychol. 1990;26(6):905910.